I think the problem with that approach, in today's market, comes when you try to hire.
Junior engineers didn't cost more than any other recent graduate, and as you could get away with a few rough edges in production for quite a while, you could put together an adequately balanced team for not much money.
The tooling was much worse, but the compliance burden was much less.
Did your friends in the .com era talk about "doing a startup" or "starting a business"? I feel like that's changed a bit, people are more calculated and cynical about how the game is to be played with respect to rounds, exit strategies and so on. The "startup" model specifically means something a bit different to "starting a business", around ambition, scale, investment model etc. All businesses need to be started, but not all businesses are startups.
I don’t think the idea of starting a business in tech these days commonly exists independently of the modern tech conception of “doing a startup.” And with the hiring part, the idea was that you’d do the work yourselves and only hire anyone at all when you couldn’t do the work anymore, or you got a big enough contract/sale/etc that you could pay them. The fact that the exit strategy/investment/initial expensive hiring messiness exists is the symptom, not the cause. What I don’t really get is the cause, and I suspect it’s cultural rather than logistical. Not doing something unless you can ramp it fast enough to start worrying about paying anyone other than the people that decided to do it is a deliberate choice. In a restaurant, it’s not. You literally can’t do it without the initial investment. My dad was a mechanical engineer in a startup making industrial paper handling equipment in the 90s— one of the first hires after the handful of founders worked themselves to design and fabricate the first machines securing them the contracts to afford more employees.
That's partly because of growth based valuations, and fast-acting network effects. The whole model is predicated on "get there fast, get the monopoly". Which you can't either with a traditional, non-scalable business (restaurants) or even a scalable industrial business (cars) - yes you get economy of scale, but they benefit relatively little from network effects of other people owning them. Goods like fashion which don't need a service ecosystem, even less so.
Total scalability & strong network effect points toward a "get the monopoly, now" business model. Which, IDK, might suit some things but definitely isn't the right fit for all things tech.
Right — I totally understand why things are the way they are using the tech industry startup model. But using that model is relatively new, and adopting it not a prerequisite to starting a tech business, even if that’s the only way people do it these days, generally. Becoming a monopoly immediately in some market is the goal of someone that wants to become the monopoly immediately in some market. It’s neither a requirement of, nor mutually exclusive of having a good idea that you want to turn into a business. That business goal might be easier in technical businesses, but it doesn’t seem to make better or more sustainable companies. “That’s the way it is done” is usually the worst reason for doing something. It doesn’t mean it’s bad, per se, but it definitely doesn’t mean it’s good.
Junior engineers didn't cost more than any other recent graduate, and as you could get away with a few rough edges in production for quite a while, you could put together an adequately balanced team for not much money.
The tooling was much worse, but the compliance burden was much less.
Did your friends in the .com era talk about "doing a startup" or "starting a business"? I feel like that's changed a bit, people are more calculated and cynical about how the game is to be played with respect to rounds, exit strategies and so on. The "startup" model specifically means something a bit different to "starting a business", around ambition, scale, investment model etc. All businesses need to be started, but not all businesses are startups.